Subscribe to get new posts from mleg.tech in your reader of choice.
Reducer bugs, state bloat, and the deep-merge fix that took a full day to trace — lessons from running multi-agent workflows on AWS Bedrock at scale.
Why fixed-size chunking quietly tanks retrieval quality, and what content-hash dedup buys you at scale.
How an eval pipeline caught stale-context regressions before they reached production, and what it cost to build.
Decoupling frontend from backend on multi-step AI processes without losing type safety or introducing polling.
What actually changes on a team's velocity and review process once AI-assisted engineering is the default, not the exception.
A decade of component-driven thinking turned out to be surprisingly good prep for orchestrating agents.
<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom"><title>mleg.tech — Michael Legemah</title><subtitle>Notes on building AI systems that ship.</subtitle><link href="https://mleg.tech/feed.xml" rel="self"/><link href="https://mleg.tech/"/><id>https://mleg.tech/</id><updated>2026-06-14T09:00:00.000Z</updated><author><name>Michael Legemah</name></author><entry><title>What Actually Breaks When You Put LangGraph Agents in Production</title><link href="https://mleg.tech/blog/langgraph-agents-in-production"/><id>https://mleg.tech/blog/langgraph-agents-in-production</id><published>2026-06-14T09:00:00.000Z</published><updated>2026-06-14T09:00:00.000Z</updated><summary>Reducer bugs, state bloat, and the deep-merge fix that took a full day to trace — lessons from running multi-agent workflows on AWS Bedrock at scale.</summary><category term="Agentic AI"/></entry><!-- 5 more entries --></feed>